R G Fiddian-Green1, S Baker. 1. Department of Surgery, University of Massachusetts Medical Center, Worcester 01655.
Abstract
OBJECTIVE: To examine the possibility that nosocomial pneumonias might be caused by the translocation of enteric bacteria and their toxins. DESIGN: Prospectively collected previous database was examined by logistic regression analysis. SETTING: University medical center. PATIENTS: Sixty-two ICU patients. MEASUREMENTS AND MAIN RESULTS: The best stand-alone predictors for nosocomial pneumonia were bleeding from stress ulceration (p less than .001), the severity of illness present (p less than .001), and intramucosal acidosis in the stomach (p = .023), a metabolic indication of mucosal ischemia. Mechanical ventilation (p = .038) and the administration of antacids/cimetidine (p = .054) were also of stand-alone predictive value, but did not significantly improve the best predictive model for nosocomial pneumonia derived from the severity of illness present and the intramucosal pH in the stomach. CONCLUSIONS: The findings are consistent with the hypothesis that ischemic mucosal injury and its associated translocation of enteric bacteria and toxins might be more important in the pathogenesis of nosocomial pneumonia in the critically ill than the aspiration of contaminated nasopharyngeal secretions.
OBJECTIVE: To examine the possibility that nosocomial pneumonias might be caused by the translocation of enteric bacteria and their toxins. DESIGN: Prospectively collected previous database was examined by logistic regression analysis. SETTING: University medical center. PATIENTS: Sixty-two ICU patients. MEASUREMENTS AND MAIN RESULTS: The best stand-alone predictors for nosocomial pneumonia were bleeding from stress ulceration (p less than .001), the severity of illness present (p less than .001), and intramucosal acidosis in the stomach (p = .023), a metabolic indication of mucosal ischemia. Mechanical ventilation (p = .038) and the administration of antacids/cimetidine (p = .054) were also of stand-alone predictive value, but did not significantly improve the best predictive model for nosocomial pneumonia derived from the severity of illness present and the intramucosal pH in the stomach. CONCLUSIONS: The findings are consistent with the hypothesis that ischemic mucosal injury and its associated translocation of enteric bacteria and toxins might be more important in the pathogenesis of nosocomial pneumonia in the critically ill than the aspiration of contaminated nasopharyngeal secretions.